Polyoxyalkylene polyether polyols are now a mainstay of the polyurethane industry, and have a myriad of other uses outside of polyurethanes as well, e.g. as surfactants, fat substitutes, and the like. While higher alkylene oxides such as butylene oxides and higher .alpha.-olefin oxides are used in the preparation of some polyether polyols, particularly those used as surfactants, the majority of polyether polyols produced today are prepared by polyoxyalkylation with ethylene oxide, propylene oxide, or mixtures thereof, in either block, random, or block random fashion. Oxyalkylation has been generally conducted in the presence of a basic metal catalyst such as sodium hydroxide, potassium hydroxide, or an alkali metal alkoxide. Such catalysts are relatively inexpensive, and readily available.
During base catalyzed oxyalkylation with propylene oxide, however, a competing rearrangement of propylene oxide into allyl alcohol during the course of the reaction continually introduces a monohydroxyl-functional molecule, which itself is capable of being oxyalkylated. Thus, as the reaction progresses, more and more monofunctional species and their oligomeric oxyalkylated products accumulate, reducing the average functionality of di- and poly-functional polyols and broadening their molecular weight distribution as well. Discussion of the mechanism of this rearrangement is discussed in BLOCK AND GRAFT POLYMERIZATION, v. 2, Ceresa Ed., John Wiley & Sons, on pages 17-21. The monol content in conventionally base catalyzed polyoxypropylene polyols is ascertained by measuring unsaturation content, for example by ASTM D-2849-69, "Testing Urethane Foam Polyol Raw Materials", and is generally in the range of 0.08 meq/g to 0.10 meq/g or higher.
For example, in the production of a 2000 Da (Dalton) equivalent weight polyoxypropylene diol, it is not uncommon for the mol percentage of unsaturated monofunctional species to approach 30-40%, lowering the theoretical functionality of 2 to a functionality in the range of 1.6 to 1.7. It is believed by some that the high monofunctionality of the polyols produced through base catalysis places a severe limitation on the ultimate molecular weight of polymers obtained therefrom due to the ability of the monofunctional species to act as chain terminators during polymerization. Moreover, the presence of the unsaturated allyl group may sometimes have deleterious effects associated with the reaction of the allylic double bond or its oxidation into a variety of oxidation products. Thus, attempts have been made to reduce the amount of monofunctional species as reflected by measuring the unsaturation content of polyether polyols.
For example, use of rubidium and cesium hydroxide as oxyalkylation catalysts in place of the normally used sodium and potassium hydroxides has been found to lower the degree of unsaturation as disclosed in U.S. Pat. No. 3,393,243. However, the decrease is somewhat modest and the catalysts are far more expensive. Use of barium and strontium hydroxides and oxides is disclosed in U.S. Pat. Nos. 5,010,187 and 5,114,619. In addition to being far more expensive than alkali metal hydroxides, both barium and strontium, particularly the latter, are relatively toxic. Thus, little if any commercialization of such processes have been made. The use of metal naphthenates, optionally in conjunction with tertiary amine co-catalysts, has been shown to be capable of producing polyoxyalkylene polyols of modest molecular weights with unsaturations as low as 0.02 to 0.04 meq/g. See, e.g., U.S. Pat. No. 4,282,387. However, these polyols still contain an appreciable monofunctional content, and have been shown to be little different their behavior in polymer systems from higher unsaturation polyols produced through conventional base catalysis.
In the decade of the 1960's, non-stoichiometric double metal cyanide catalysts, i.e. the glyme adduct of zinc hexacyanocobaltate, were shown to be efficient catalysts for polyoxyalkylations and various other reactions as well. In addition to possessing relatively high rates of reaction, the double metal cyanide catalysts were found to be capable of producing polyols with relatively low unsaturation content, with unsaturations in the range of 0.018 to 0.020 easily obtainable. However, due to the greater cost of these catalysts, coupled with the difficulty of removing them from the finished polyol product, commercialization of such systems did not materialize. Interest in DMC catalysts resurfaced in the 1980's, and improvements in the catalytic activity coupled with new and improved methods of removal from the finished product resulted in commercialization of DMC catalyzed polyol production for a short period of time. The polyols produced by these improved DMC catalysts exhibited unsaturation in the range of 0.015 to 0.018 meq/g.
Recent further improvements in the activity of DMC catalysts by the ARCO Chemical Company has once again resulted in commercialization of DMC catalyzed polyoxyalkylene polyols under the tradename ACCLAIM.TM. polyols. The catalytic activity of the new DMC catalysts has been improved to such an extent that frequently the rate of polyoxyalkylation is limited by the ability to transfer heat from the polymerization reactor rather than by the activity of the catalyst. The decreased processing time increases the cost/benefit ratio of the catalysts, encouraging their commercial use.
The availability of ultra-low unsaturation polyoxyalkylene polyols has not proven to be the panacea expected. The polyols have not proven to be drop-in replacements for conventionally catalyzed polyether polyols, and in fact, polyether polyols produced by DMC catalysts having unsaturation in the range of 0.003 to 0.010 meq/g have surprisingly been found to be quantitatively different from other, "low" unsaturation polyoxyalkylene polyols, even those produced by prior generation DMC catalysis. Although a portion of the differences between ultra-low unsaturation polyether polyols, low unsaturation polyols, and conventional polyoxyalkylene polyols can be attributed to the differences in functionality, molecular weight distribution, and lack of monofunctional species, the manner in which each of these properties affects both polyurethane formulation and the polyurethanes obtained therefrom is not fully understood. In many cases, for example, improved polyurethane products could indeed be produced, but only by unusual and non-obvious changes in formulation and processing parameters. The anomalous behavior of ultra-low unsaturation polyoxyalkylene polyols and the reasons therefore are still being investigated. One example of this anomalous behavior is the collapse of polyurethane foam systems which employ high secondary hydroxyl content ultra-low unsaturation polyoxypropylene polyols, while similar systems employing conventional polyols produced good foams.
It has been recently discovered that polyoxypropylations with some double metal cyanide catalyst systems, particularly those capable of producing ultra-low unsaturation polyether polyols, also produce a very small but significant quantity of very high molecular weight product. The existence of this very high molecular weight "tail" was not expected, as double metal cyanide catalysts are known to produce polyether polyols of very low polydispersity, with polydispersities on the order of 1.07 to 1.20 being routinely achieved. This polydispersity is far lower than those obtainable with alkali metal catalysis and other methods of catalyzing the oxyalkylation, and gel permeation chromatography shows a relatively tight and narrow distribution of molecular weights.
Upon careful analysis of larger quantities of polyol, concentrating on the portion eluting significantly prior to the main product peak, a very high molecular weight component was surprisingly discovered. Careful analysis of this high molecular weight "tail" indicates that it is composed mostly of polyoxypropylene polyols having molecular weights in excess of 100,000 Da. Once being appraised of the existence of the high molecular weight tail, the reasons for its production may be hypothesized. Without wishing to be bound to any particular theory, Applicants believe that the non-stoichiometric double metal cyanide catalysts contain a very minor portion of catalytic sites for which the transfer coefficient is exceptionally small. While the vast majority of catalytic sites exhibit rapid substrate transfer, resulting in a very narrow and tight molecular weight band, a small fraction of the catalytic sites may exhibit virtually no transfer whatsoever, thus producing at those sites a higher and higher molecular weight product. Applicants believe that this high molecular weight tail may be one facet of the explanation of the anomalous behavior of DMC catalyzed ultra-low unsaturation polyether polyols in some applications.
One example is the production of polyurethane foam from high secondary hydroxyl polyoxypropylene polyols. It is known that the hydrophilicity and hydrophobicity of polyethers is affected by molecular weight. Further, it is believed that the high molecular weight species is virtually all polyoxypropylene homopolymer. Thus, while ethylene glycol, oligomeric ethylene glycols, and even high molecular weight polyoxyethylene glycols are all hydrophilic to some degree, polyoxypropylene glycols are hydrophilic only up to a molecular weight of approximately 500 Da, following which they become increasingly hydrophobic. This phenomenon has been utilized in the preparation of polyoxyethylene/polyoxypropylene block copolymers useful as nonionic block surfactants.
In the production of polyurethane foam, the foam chemistry is very critical. For example, acceptable polyurethane foams are rarely obtained when incompatible polyols are utilized in the foam formulation. An incompatible polyols is one which is insoluble or of limited solubility in the unreacted or partially reacted foam forming ingredients. This is one reason polyethylene glycols are rarely used in foam formulations except in most minor amounts, as the polyoxyethylene polyols are generally of very limited solubility in the reactive components. On the other hand, polyoxypropylene polyols of modest molecular weight tend to be compatible in this respect. During the condensation polymerization which takes place during foam formation, the growing polyurethane and/or polyurethane/urea polymers increasingly incorporate relatively polar urethane and urea groups, thus altering the hydrophile/lipophile balance of the growing polymer chain. It is believed that foam collapse occurs in foam formulations employing high secondary hydroxyl, ultra-low unsaturation DMC catalyzed polyoxypropylene polyols having a high molecular weight tail, because the high molecular weight tail, being exceptionally hydrophobic, tends to phase out, or separate from the growing polymer matrix, disrupting cell walls and eventually causing foam collapse.
Elimination of the high molecular weight tail from polyoxyalkylene polyols containing such a tail by removing the latter is virtually impossible, as in general, the polyoxyalkylene polyol products are of far too high a molecular weight for efficient distillation, even with methods such as falling film evaporation or wiped film evaporation, and moreover, such evaporative processes are relatively expensive, adding unwanted cost to the polyether product. If a method of precipitating the high molecular weight tail could be found, the high molecular weight portion could be filtered out. However, filtration of relatively viscous polyols is a long and expensive process. Again, the cost benefit ratio would dictate against the use of such a method even if such a method were available.
It would be desirable to produce low and ultra-low unsaturation, high molecular weight polyoxyalkylene polyols with high secondary hydroxyl content wherein the phase-out of the high molecular weight tail during polyurethane formation is markedly decreased or eliminated. It would further be desirable to produce a polyether polyol product wherein the high molecular weight tail is less hydrophobic than a polyoxypropylene homopolymer.